The big exception to these generalizations was Stalin. Liberals and conservatives strongly disagreed over the plausibility of the antecedent as well as over the conditional linkage for the Stalinism counterfactual. Conservatives had a harder time than liberals imagining that the Soviet Communist Party could have purged, or would have wanted to purge, Stalin in the early 1930s. From a conservative perspective, which views Stalinism (not just Stalin) as the natural next step of Leninism, the deletion-of-Stalin counterfactual violates the minimal-rewrite rule. But this counterfactual may well pass the minimal-rewrite test for those with a more liberal perspective on Soviet polity.
Liberals and conservatives also disagreed on what would have happened if Stalin had been deposed. Like most historical counterfactuals, this one does not spell out the complex connecting principles necessary for bridging the logical gap between antecedent and consequent. To hold the counterfactual together, it is necessary to posit that advocates of Gorbachev-style socialism in the Communist Party of the Soviet Union (CPSU) would have seized the reins of power and guided the Soviet state toward social democracy. Conservatives regard such arguments as fanciful.
Taken as a whole, these data are open to two rival interpretations. One hypothesis asserts that those on the left view history as more fluid, contingent, and indeterminate than those on the right. The other asserts that liberals and conservatives do not have fundamentally different philosophies of history but there is something special about the Soviet Union that motivates wistful perceptions of lost possibilities on the left and angry accusations of inevitable repression and expansion on the right. If we could identify a state that excites fear and loathing on the left comparable to that once excited by the Soviet Union on the right, we would observe a sign reversal of the correlation coefficients between ideological sympathies and counterfactual beliefs. South Africa was the ideal case for teasing these hypotheses apart.
Demise of White-minority Rule in South Africa
Competing Schemas. Observers on the left now leaned toward essentialism. The incorrigibly racist, white-minority regime would cede power only under enormous pressure. Observers on the right favored a more pluralistic view of politics inside Pretoria. They sensed that “verligte” or enlightened factions were eager to enter into flexible power-sharing arrangements.
Counterfactual Probes. This disagreement was as close to a mirror image of the controversy over the Soviet Union as nature was going to provide. We therefore hypothesized a reversal in relative openness to close-call counterfactuals: the right would now embrace counterfactuals that assign a key role to political personalities within the regime (e.g., “If no de Klerk, then continued impasse”), and the left would now embrace counterfactuals that assign a key role to external pressure (e.g., “If no Western sanctions, then continued minority rule”). The operative principle is dissonance reduction: the more we hate a regime, the more repugnant it becomes to attribute anything good to redemptive dispositions of the regime (such as a capacity for self-correction).
Findings. Consistent with the two-stage model of counterfactual inference, political ideology was again an anemic predictor of the mutability of historical antecedents but a robust predictor of antecedent-consequent linkages. Conservatives assigned more credit to de Klerk and to the collapse of Soviet-style Communism, whereas liberals gave more credit to Western sanctions.
The debate over the impact of sanctions on South Africa does indeed closely mirror the debate over the impact of Reagan’s defense buildup on the Soviet Union. Paraphrasing sentiments some Western liberals attributed to the Soviet elite at the dawn of the Gorbachev period, one conservative argued that the momentum for change inside South Africa had become irresistible because white elites had concluded from the township revolts and demographic trends that they “could not go on living this way.” Another conservative argued that credit for ending white-minority rule should go to Reagan, whose policies precipitated the implosion of Soviet Communism, allowing de Klerk to convince his followers that negotiating with the ANC was not tantamount to surrender to the Kremlin. We thus come full circle. The observer refuted the argument that “even if you conservatives were right about the Soviet Union, you were wrong about South Africa” by arguing that “it was because we conservatives were right about the Soviet Union that we were also right about South Africa.”
Discussion. Conservatives’ openness to the de Klerk counterfactual in the South African case parallels liberals’ openness to the Stalin, Malenkov, and Gorbachev counterfactuals in the Soviet case; liberals’ skepticism toward the Reagan-pressure counterfactual in the Soviet case parallels conservatives’ skepticism toward the economic sanctions counterfactual in the South African case. These data undermine the sweeping claim that liberals subscribe to a more contingent view of philosophy of history than conservatives. Much hinges on whose “policy ox” is being gored.
Taken together, the two studies show that beliefs about specific counter-factual possibilities were rather tightly coupled to overall ideological outlooks. But neither study was well equipped to test the cognitive-style hypothesis that hedgehogs are more likely than foxes to reject close-call counterfactuals that undercut their pet theories. And both studies were conducted when the policy debates had only recently been rendered moot (1992 in the Soviet case, 1995 in the South African case), leaving open the question of what happens when scholars contemplate counterfactuals that undo events further removed from current controversies. Does temporal distance reduce the iron grip of our preconceptions on our judgments of what could have been? The next four studies address these issues.
Rerouting History at Earlier Choice Points
UNMAKING THE WEST
Competing Schemas. Historians have long puzzled over how a small number of Europeans, and their colonial offshoots, came between 1400 and 1700 C.E. to exert such disproportionate influence around the globe. The resulting debate has polarized scholars. In one camp are determinists who view Western geopolitical ascendancy as having been inevitable for a long time. Western culture possessed critical advantages in the Darwinian struggle for societal survival: more deeply rooted traditions of private property and individual rights, a religion that encouraged achievement in this world, and a fractious multistate system that prevented any single power from dominating all others and bringing all innovation to a grinding halt whenever reactionary whims struck the ruling elite. In the other camp are the radical antideterminists who believe, to adapt Gould’s famous thought experiment, that if we were to rerun world history repeatedly from the same conditions that prevailed as recently as 1500 C.E., European dominance would be a rare outcome. The European achievement was a precarious one that can be easily unraveled.
Counterfactual Probes. Antideterminists have generated a long list of close-call counterfactuals designed to puncture “Eurocentric triumphalism”: South Asia, East Africa, and perhaps the Americas might have been colonized by an invincible Chinese armada in the fifteenth century if there had been more support in the imperial court for innovation and expansion; Europe might have been conquered and Islamicized in the eighth century if the Moors had cared to launch a serious invasion of southern France and Italy; and European civilization might have been devastated by Mongol armies in the thirteenth century but for the fortuitous timing of Genghis Khan’s death.
Findings. Table 5.2 shows that the more experts embraced deterministic explanations for Western dominance, the more dismissive they were of counterfactuals that implied that the West was just luckier than the Rest and the more prone they were to reject counterfactuals that implied that other civilizations could have blocked Western hegemony or achieved dominance themselves. The hypothesized interaction also emerged: the power of preconceptions to predict reactions to counterfactuals was greater among the hedgehogs who—as one might expect by now—struggled harder to squeeze history into their ideological frameworks. By contrast, foxes were more tolerant of counterfactuals that poked holes in their ideological frameworks. This cognit
ive-style-by-ideological-worldview interaction also held in the final three studies, which explored reactions to rewrites of twentieth-century history.
THE OUTBREAK OF WORLD WAR I
Competing Schemas. Some scholars believe that war among the Great Powers of Europe in the early twentieth-century was inevitable. This thesis is often grounded in causal arguments that stress the inherent instability of multiethnic empires and multipolar balances of power, as well as the “cult of the offensive” (the widespread perception among general staffs that the side that struck first would gain a decisive advantage).
TABLE 5.2
Predicting Resistance to Close-call Counterfactuals
Note: This table presents the results of multiple regressions that treat resistance to close-call counterfactuals as the dependent variable and treat as independent variables experts’ commitment to theoretical schools of thought (neorealist balancing, robustness of nuclear deterrence, adaptive advantage of West), cognitive style (need for closure/reflected version of hedgehog-fox scale), and a cross-product term for capturing whether resistance is greatest when both theoretical commitment and need for closure are highest.
Counterfactual Probes. The more experts endorse these “macro” causal arguments, the more ill-disposed they should be toward counterfactuals that imply that war could have been avoided by undoing one of the bizarre coincidences preceding the assassination of Archduke Ferdinand or by permitting minor alterations of the content or timing of diplomatic messages exchanged among the Great Powers in the six weeks preceding the outbreak of war.
THE OUTCOMES OF WORLD WARS I AND II
Competing Schemas. Neorealist balancing, a profoundly influential framework in world politics, asserts that when one state threatens to dominate the international system, other states coalesce to preserve the balance of power.10 It is no accident that would-be world conquerors such as Philip II, Napoleon, and Hitler failed: their failures were predetermined by this fundamental law of world politics.
Counterfactual Probes. The more experts endorse neorealist balancing, the more ill-disposed they should be to close-call counterfactuals that imply that the Germans could easily have emerged victorious in either of the two world wars and achieved at least continental hegemony if they had made better strategic-military decisions at key junctures.
WHY THE COLD WAR NEVER GOT “HOT”
Competing Schemas. Some scholars believe in the robustness of nuclear deterrence and mutual assured destruction: rational actors do not commit suicide.11 When these scholars look back on the cold war, they find it hard to imagine that crises could have escalated out of control (just as they have a hard time getting agitated about future dangers of nuclear proliferation).
Counterfactual Probes. These scholars should be dismissive of close-call counterfactuals in which the United States and USSR slip into nuclear war at various junctures in the cold war (e.g., if Kennedy had heeded the advice of his hawkish advisers and launched air strikes against Soviet missile sites in Cuba in October 1962, or if Eisenhower had followed through on his threat to use nuclear weapons to break the stalemate in the Korean War).
Aggregated Findings. In all three twentieth-century contexts, the more committed scholars were to a generalization undercut by a counter-factual, the more dismissive they were of that counterfactual. And in all three contexts, hedgehogs were especially likely to dismiss counterfactuals that undercut their theoretical commitments. As table 5.2 underscores, counterfactuals were viewed as a nuisance at best, and a threat at worst, to analysts on the prowl for ways of achieving explanatory closure by assimilating past events into favorite theories of history.
Also, again, we observe far tighter links between theoretical orientations to world politics and the more historically transportable belief system defenses (challenging connecting principles and generating second-order counterfactuals) than we do with the most historically rooted defense (challenging the mutability of the antecedent). There is no reason why one’s position on the macro causes of war should predict whether one believes the assassination of Archduke Ferdinand could have been thwarted if his carriage driver had a better sense of direction or why one’s position on the robustness of nuclear deterrence should predict whether one believes Stalin could have survived his cerebral hemorrhage.
Curiously though, the strategy of resisting close-call counterfactuals by challenging the mutability of the antecedent was not completely decoupled from abstract beliefs. Ideological orientations predicted significant (8 percent to 12 percent) proportions of the variance in judgments of the mutability of antecedents. These results suggest that even the most apolitical facts—the reconnaissance capabilities of U-2 aircraft on partly cloudy days or the cerebrovascular health of an aging dictator—can quickly be politicized as soon as rival schools of thought discover an advantage in showing a downstream outcome to be either easy or hard to undo.
Taken in their entirety, the results might appear at odds with those in chapter 4 in which hedgehogs endorsed more close-call counterfactuals than did foxes. Context, however, matters. In chapter 4, hedgehogs showed more interest in only those close-call counterfactuals that protected their forecasts from disconfirmation, that let them save face by invoking the “I was almost right” defense. The contradiction vanishes when we consider how hedgehogs could most efficiently achieve the same face-saving goal in settings, like those here in chapter 5, in which close-call counterfactuals undermine their favorite deterministic accounts of the past. Protecting one’s belief system now requires invoking the “I was not almost wrong” defense, demonstrating that, although it might look easy to derail a historical process, on close inspection it is remarkably difficult: as soon as one cuts off one pathway to the observed outcome, other pathways arise, hydralike, in second-order counterfactuals. There is nothing stylistically inconsistent in rejecting close-call counterfactuals that challenge one’s preferred explanations of the past and embracing close calls that buffer one’s expectations about the future from refutation.
ASSESSING DOUBLE STANDARDS IN SETTING STANDARDS OF EVIDENCE AND PROOF
These data raise a worrisome question: What is to stop politically motivated observers from positing counterfactuals that justify whatever causal assertions they find it expedient to make? Tetlock and Belkin answer this question by specifying criteria for winnowing out specious counterfactuals.12 They argue that, although counterfactual claims are strictly speaking about empirically inaccessible possible worlds (no one can hop into a time machine, undo key events, and document what would have happened), it is often possible to test the implications of such claims in this world. Indeed, there is a voluminous literature on the logical, statistical and historical criteria that scholars should use in judging counterfactuals.13 Most of us suspect that some counterfactuals are more compelling than others, and this literature suggests that there are good grounds for holding to this suspicion. But there is a cognitive catch: to prevent speculation from sliding into the solipsistic abyss, experts must be willing to change their minds about possible worlds in response to real-world evidence. As we shall now see, many are reluctant.
Let us revisit the sharply contested counterfactual “If the Communist Party of the Soviet Union had deposed Stalin in the late 1920s, the USSR would have moved toward a kinder, gentler form of Communism fifty years earlier.” As a thought experiment, suppose that historical sleuths in the Kremlin archives claim to discover documents that reveal rising resistance to Stalin in the late 1920s and that, given the chance, the most likely successors would have constructed a kinder, gentler Communism. How should experts respond? It seems a trifle dogmatic to refuse even to consider changing one’s mind. But such a response might be justifiable if overwhelming evidence from other credible sources pointed to the contrary conclusion. Many scientists justify their dismissal of evidence for “far-out claims” such as extrasensory perception on the ground that such findings violate too many well-established physical and biological laws. In Bayesian terms, it seem
s presumptuous for nonexperts to tell experts how “diagnostic” particular evidence is with respect to particular causal hypotheses.
It is possible, however, to design a better mousetrap for documenting the impact of theory-driven thinking about counterfactual history. Imagine that we transform our thought experiment into an actual experiment that holds evidence constant—say, documents recently discovered in Kremlin archives—but manipulates findings—say, whether the documents contain revelations favorable either to those who view Stalinism as an aberration or to those who view it as a natural outgrowth of Leninism. Insofar as observers deem evidence compelling only when it reinforces their prior beliefs, the experiment would reveal a disturbing double standard in judgments of the probative value of evidence. To the degree that scholars keep two sets of books for scoring knowledge claims—one set for consonant claims and the other for dissonant ones—the risk grows that their beliefs about historical causality will ossify into brittle tautologies in which they alternate between invoking ideological preconceptions to justify their claims about what could have been and invoking claims about what could have been to justify their preconceptions.
Expert Political Judgment Page 27
